Utility of mitochondrial‐encoded cytochrome c oxidase I gene for phylogenetic analysis and species identification of the planktonic diatom genus Skeletonema

Abstract

SUMMARY Small subunit (SSU) and large subunit (LSU) rDNA sequences have been commonly used to delineate the taxonomy and biogeography of the planktonic diatom genus Skeletonema, but the genes occur as multiple copies and are therefore not suitable for barcoding purposes. Here, we analyzed phylogenetic relationships of Skeletonema using the mitochondrial‐encoded cytochrome c oxidase I gene (cox1), as well as partial LSU rDNA (D1–D3) and SSU rDNA, to identify the factors that define species and to evaluate the utility of these three markers for this taxon. Twelve Skeletonema species were divided into six clades, I–VI, each of which comprised the same species by the three markers: clades I (S. japonicum, S. grethae, S. pseudocostatum, and S. tropicum), II (S. menzelii), III (S. dohrnii and S. marinoi), IV (S. costatum, S. potamos, and S. subsalsum), V (S. grevillei), and VI (S. ardens). However, the branching order among these clades was incongruent among the markers. In clade III, six S. marinoi strains had identical cox1 sequences. These S. marinoi strains branched along with S. dohrnii, except for strains from the Gulf of Naples, with high support in cox1. Species delimitation between S. dohrnii and S. marinoi was therefore not supported. In clade IV, S. costatum and S. subsalsum were robustly clustered, with S. potamos as a sister clade in the cox1 tree, not in the LSU and SSU trees. In clade II, cox1 also confirmed that S. menzelii includes three subclades potentially distinguishable from each other by morphological features. Cox1 proved to be the most useful marker for the identification of Skeletonema species because it gave a tree with highly supported clades, has sufficient variation within and among species, encodes a protein in a single copy, and requires relatively few primers.

Comparison of environmental DNA and bulk‐sample metabarcoding using highly degenerate cytochrome c oxidase I primers

“…However, we acknowledge that COI, as all genetic markers, has important limitations. Although it has been found to be a suitable marker for several non‐metazoan taxa (diatoms: Yamada et al, ; Oomycota: Robideau et al, ; Amoeboaza: Nassonova et al, , Rhodophyta: Saunders & McDevit, ), care should be taken when interpreting data and further research is needed in order to fully understand the usability of this marker for non‐metazoans. This applies especially to Stramenopiles, as open questions remain regarding the identity of species in this highly diverse non‐monophyletic taxon and the resolution that can be achieved by using COI (Pawlowski et al, ), and COI references are scarce compared to metazoan taxa.…”

“…The COI gene is not frequently used for metabarcoding of non‐metazoan taxa due to unsolved questions regarding the ability of the marker to distinguish microbial taxa (Pawlowski et al, ) and the lack of reference databases. However, studies have shown that COI might be a good marker for some non‐metazoan taxa (diatoms: Yamada et al, ; Oomycota: Robideau et al, ; Amoeboaza: Nassonova, Smirnov, Fahrni, & Pawlowski, ; Rhodophyta: Saunders & McDevit, ). As COI is the standard marker for animals (Ratnasingham & Hebert, ), highly degenerate COI primers also amplifying non‐metazoan taxa might bridge a gap between COI metabarcoding of animals and metabarcoding of non‐metazoan taxa and could provide useful complementary information on non‐metazoan taxa for ecological studies and biodiversity assessments.…”

A polyphasic approach to the delimitation of diatom species: a case study for the genusPinnularia(Bacillariophyta)

“…) and Skeletonema (Yamada et al. ) and the investigation of diatom species diversity in relation to DNA barcoding (Trobajo et al. , Hamsher et al.…”

Effect of cell size on growth rates of two Skeletonema species and their competitve interactions

“…Moreover, Yamada et al. () reported that species delineation between S. marinoi and S. dohrnii was not supported by molecular analysis using the small subunit rDNA, LSU rDNA, or the mitochondrial‐encoded cytochrome c oxidase I gene (COX1). In this study, we tentatively identified our strains, MD1 and MD5 as S. dohrnii because strains were in the S. dohrnii cluster based on LSU sequence.…”

“…Among the species, Skeletonema marinoi and Skeletonema dohrnii are morphologically indistinguishable, even using scanning electron microscopy (e.g., Ellegaard et al 2008, Kooistra et al 2008. Moreover, Yamada et al (2017) reported that species delineation between S. marinoi and S. dohrnii was not supported by molecular analysis using the small subunit rDNA, LSU rDNA, or the mitochondrial-encoded cytochrome c oxidase I gene (COX1). In this study, we tentatively identified our strains, MD1 and MD5 as S. dohrnii because strains were in the S. dohrnii cluster based on LSU sequence.…”